# BLOCKLORE — full documentation > Source pages: https://blocklords.ai/en/docs and https://blocklords.ai/en/blog. Operated by Metaking Studios OU. > Index: https://blocklords.ai/llms.txt ## What is BLOCKLORE Infinite canvas + node-based workflows + AI image / video generation + video analysis and storyboarding. Upload assets, connect, click to generate—every step in between lands on the canvas, ready to adjust, reuse, and export. ## Product capabilities Multimodal AI canvas: image, video, and audio generation nodes; storyboard and director stage; template library; community publish; credit-based subscriptions and packs. ### Images Text-to-image and reference-driven generation on canvas nodes, plus editing tools such as background removal and relight where the model supports them. ### Video Text-to-video, image-to-video, first/last-frame control, reference video, and extend or edit passes — each as a node you can rerun. ### Audio Speech, music and voice nodes, with library pickers, so a cut arrives with sound instead of waiting on another tool. ### Video editor A multi-track timeline for stitching, layering, subtitles and transitions, exporting to your computer or back onto the canvas. ## How BLOCKLORE differs from a stack of separate tools ### Infinite canvas React Flow node workflows with type-checked links ### Unified model layer Image / video / audio generation, all through BLOCKLORE ### Cloud assets Results become assets automatically, stored in OSS, ready to reuse and export ## Pricing FAQ ### Credit validity rules 1. Membership credits are issued monthly. Each grant is valid until its listed expiry date; quarterly and annual plans continue to receive credits monthly until membership ends. 2. Personal fixed general packs: paid and bundled bonus credits are valid for 730 days. Custom top-ups: purchased credits are valid for 730 days and bonus credits for one calendar month. Team fixed packs: paid credits are valid for 730 days and bonus credits for one calendar month. 3. Model-exclusive credits follow the validity shown on the pack card. 4. Daily login credits are valid only on the day they are granted. ### Membership & credits refund policy Membership and credits are virtual digital goods and take effect immediately after purchase; consumed credits cannot be refunded or transferred. You can cancel auto-renewal anytime in subscription management — no further charges after cancellation. For duplicate charges or system issues, please contact support and we'll resolve it promptly. ### Credit refund on failure If a generation task fails, the deducted credits are automatically returned to the original account, entry by entry, following the original charge. Portions that cannot be returned because the original credit lot has expired are compensated with equal gifted credits (valid for 24 hours). On cancel: queued tasks that have not yet called a model provider are fully refunded; once a provider call has been claimed, cancel is non-refundable. Contact support if a refund is delayed or missing. ### Credit consumption order Default consumption order: gift credits > model-exclusive credits > membership credits > purchased general credits. Within the same type, credits expiring sooner are consumed first. ### How to get more credits - Subscribe or upgrade: membership credits are issued monthly. - Buy fixed or custom general-credit packs; check each pack's validity before purchase. - Buy model-exclusive packs for supported model series. - Claim daily login credits. - Join platform contests or redeem codes for credit rewards. ### Invoices & contact For invoices or business cooperation, please reach us via "Contact us" at the bottom of the page and we'll get back to you promptly. ### Cancellation & expiry After cancellation, your subscription no longer renews automatically. Granted credits remain usable until their own expiry dates, and quarterly or annual plans continue to receive monthly grants until membership ends. The account then returns to the Free tier; unexpired purchased and model-exclusive credits are unaffected. ### Contact support Reach us via "Contact us" at the bottom of the page or the Help Center — we'll reply as soon as possible. ## Getting started BLOCKLORE is an AI creation tool built around a single canvas. On an infinite, freely zoomable surface you place nodes — images, video, audio, text, scripts, timeline edits — wire them together, and drive image, video, audio and voice generation until you have a finished piece. This handbook is written against the real interface. Pick a chapter on the left; each page shows one chapter only, and every step is something you can follow along. Work through this chapter once and you will go from signing up to your first generated image. ### Sign up and sign in - Click Sign up / Sign in at the top-right of the nav. Email code, email password, Google / Discord / Telegram (shown per site config) and wallet-signature sign-in are all supported. - While signed out, any action that needs an account (new canvas, fork a project, publish, and so on) pops up the sign-in dialog instead of forcing a page jump; only opening the canvas editor itself requires signing in first. - New accounts usually come with starter credits, shown in the top bar or the account menu. Generation is billed, and the starter credits are enough to try a few images. ### Open a canvas There are three ways to create a canvas; any one of them drops you straight into the editor: - On the home Workspace block, click the New card, or click any of your Recent projects to open it; - Open the Workspace page (nav Workspace, i.e. /dashboard) and click New canvas under the Private or Team tab; - Open a template (/templates) or the Templates tab in Workspace and click Fork project to create an editable canvas from that template. (Image: Home: promo carousel, the Workspace New card and recent projects) ### Your first win: text to image Getting your first image on an empty canvas takes four steps: - Double-click an empty spot on the canvas (long-press ~0.5s on touch), choose Image in the Add node menu to create an image node; - Select that node — a generator bar appears at the bottom — and type the picture you want in the prompt box, e.g. "a lighthouse on a seaside cliff at dusk, cinematic"; - Pick a model in the generator bar; the credits this run will cost show live next to the send button; - When it looks right, click send (or press ⌘/Ctrl + Enter). After a moment the result lands back on the node. Your edits save automatically; the top bar shows Saved / Saving and the like, so there is no manual save. (Image: Select an image node and the bottom generator bar appears: prompt, model, estimate, send) ### The pieces of the interface to know Once signed in, the desktop top nav reads, left to right: Home → Workspace → Community → AI Plays → Blog, with a Pricing link and the avatar menu on the right. - Home: the promo carousel, your Recent projects, a community masonry, and entry points to Blog and templates; - Workspace (/dashboard): your studio, split into Private, Team and Templates tabs; - Community: browse works others have published; once signed in you can favorite, publish, and see your own history; - Avatar menu: language switch, dark/light theme, the Tutorials & Docs entry, account settings, sign out. On mobile, Templates, Teams, the generation queue, Docs and Settings also fold into a side drawer. > Always check the credit estimate before you submit · The canvas saves automatically — there is no save button · Starter credits are enough to run your first image ## Account and workspace This chapter covers the pages you will use every day once signed in: the top nav, the workspace, the asset library, and the watermark switch in settings that directly affects your canvas results. ### Top navigation The top nav is the main entry point for the whole site. On desktop it runs Home, Workspace, Community, AI Plays, Blog, with Pricing and the avatar menu on the right; signed-out visitors get the language switcher in the top bar, and once signed in the language and theme controls move into the avatar menu. - The avatar menu switches interface language and dark/light theme, and opens Tutorials & Docs, account settings and sign out; - On mobile, Templates, Teams, the generation queue, Docs and Settings fold into a right-hand drawer, opened from the avatar or menu button. ### Workspace The Workspace page (/dashboard) is your studio, split into three tabs: - Private: your own canvases — organize into folders, rename, set covers, duplicate, move, delete; - Team: canvases that belong to a team; you first need to join or create a team (see Teams and settings); - Templates: a market of official and community templates — click one to spin up a new canvas from it. There is also an Import button at the top, for bringing an external canvas in (see the import note in Assets and history). (Image: Workspace: usage overview, generation queue, and the Private / Team / Templates tabs) ### Asset library The asset library (from the avatar menu on desktop, or directly at /assets; in the drawer on mobile) gathers everything you have generated: - History: filter by image / video / audio / document, sort, favorite, download and delete every asset you have generated; - Saved: the collection you have favorited — view by category, rename, recategorize, download, delete. ### Settings and watermark Open the avatar menu and go to Settings; the one thing there that directly touches your canvas results is the watermark switch: turn it on and your generated images and videos get a small watermark (position, opacity, text or a custom logo are configurable) so authorship is traceable; off means no watermark. The other sections of Settings (profile, security, account binding, credits, notifications, compliance) are not tied to canvas work and are out of scope here. Language and theme are not in Settings — they live in the avatar menu dropdown. (Image: Settings: the watermark switch is the one tied to your canvas results) > Workspace has Private / Team / Templates tabs · The library only views assets; put them back via the node menu · Watermark is in Settings; language/theme are in the avatar menu ## The infinite canvas The canvas is the main stage of BLOCKLORE: an infinitely pannable, zoomable board where nodes sit and connections show how data flows. This chapter introduces the canvas interface and how to move, zoom and pan it. ### Reading the canvas interface Once you open a canvas (/canvas/your-canvas-id), the interface breaks down into a few parts: - Top bar: on the left, the canvas title (click to rename) and the save-status hint (Saved / Saving / Unsaved, etc.); on the right, Share, Publish and other entries. A team canvas also shows an "N online" badge centered at the top; - The canvas body: an infinitely pannable, zoomable board, nodes on it, connections showing data flow; - The left vertical toolbar: entries such as the Toolbox (official templates + your saved workflows) and Shortcuts (opens the cheat-sheet panel); on narrow screens these fold into a More tools or Help menu. (Image: Empty canvas: top bar, left toolbar, the double-click hint and four quick-start cards) ### Quick-start cards A new empty canvas shows a "double-click / freely add a node" hint in the center, plus four Quick-start cards — Story script, Character three-views, First-frame image to video, Audio to video. Click any one to drop in a preset set of example nodes and connections as a starting template. The Story script card is the entry point for creating a Script node — the Script node does not appear in the regular Add node menu, so this card is the only way in. ### Move, zoom and pan - Marquee / hand toggle: press V for marquee mode (drag a box to select nodes), H for hand mode (drag the canvas itself); - Zoom: pinch on a trackpad, or hold ⌘/Ctrl and scroll the mouse wheel; ⌘/Ctrl + = to zoom in, ⌘/Ctrl + - to zoom out; - Fit view: ⌘/Ctrl + 0 auto-zooms and centers everything; - Pan: hold space and drag, or drag with the middle / right mouse button. > V marquee / H hand; ⌘/Ctrl+0 fits everything · The empty-canvas quick-start cards are the fastest way in · The Script node exists only via the Story script card ## Adding nodes Nodes are the building blocks on the canvas. Asset nodes hold what you upload; generation nodes drive AI output. This chapter covers how to add nodes and what node types exist. ### Ways to add a node There are many entry points, suited to different moments: - Double-click empty space / long-press on touch: the Add node menu pops up; pick a type to create; - Right-click empty space: the context menu has Add node, plus Upload, My assets, Undo/Redo, Paste; - Click the "+" beside a node (generate downstream): lists nodes you can create from this node's output type (a video's "+" expands Extend video, Edit video, etc.); - Click the "+" beside a node (generate upstream): from the input port, create an upstream node of the type you need and auto-connect it; - Drag a connection off a port and drop on empty space: the same candidate-type menu appears where you drop, and the node is created and connected; - Drag a local file onto the canvas: drop an image/video/audio/document to create the matching asset node and upload it; - Paste (⌘/Ctrl + V): an external image on the clipboard pastes as a new image asset node; a node copied inside the app pastes as a node duplicate. (Image: Add-node menu: Text, Image, Video, Video editor, Director Stage, Audio, plus Upload and Choose from history) ### Node types at a glance These are the main node types you can create directly and actually use for generation: - Image / video / audio / document assets: hold what you upload, as inputs for later generation (create via upload / drag / paste); - Image generation: text to image; connect an image in and it switches to image-to-image (Add node menu, Image); - Video generation: text to video, and also first/last-frame and reference-to-video modes (Add node menu, Video); - Image-to-video / extend / edit video: created from the "+" beside an image or video node; - Text to speech, speech to text, prompt from image: created from the Add node menu or a neighbor's "+"; - Text generation: AI continuation or rewriting (Add node menu, Text); - Video editor: a timeline editing container for multiple clips, opening a standalone editor (menu Video editor, marked Beta); - Director Stage: a 3D stage for placing cameras and characters and screenshotting a reference image (menu Director Stage, marked NEW); - Script: breaks a text idea into a storyboard table (created from the Story script quick-start card on an empty canvas). ### A note on billing Asset nodes are free; only generation / analysis nodes consume credits. Every paid generation shows its estimated credit cost before you submit, and you confirm before it runs. > Double-click, right-click, or a node's "+" all add nodes · Asset nodes are free; only generation/analysis is billed · The Script node isn't in the regular menu — use the quick-start card ## The generator bar Select a node that can generate and a generator bar appears at the bottom of the canvas. This is the console for one run: write the prompt, pick the model, tune parameters, read the estimate, hit send. ### What the generator bar contains Roughly, top to bottom, the bar has: - The prompt box: describe the image / video / voice you want; type @ to reference connected nodes or items from your saved library (see References and @ mentions); - Model select: models differ in capability and price; you always see the model's display name; with none picked, the default model for that task type is used; - The parameter area: resolution, aspect, duration, count and so on, varying by model and task type; only parameters the model truly supports are shown; - The reference strip: image-to-image, image-to-video and the like show thumbnails of the connected image / video / audio; - The send button: submits the run, with the estimated credit cost shown live beside it. (Image: The generator bar: prompt box, model select, parameters, reference strip and send) ### Estimate and billing Every time you change the prompt, model or a parameter, the bar re-quotes the server and refreshes the estimate, so the price you see and the basis you are actually charged on are identical. - The estimate shows as one of "— / N credits / up to N credits": models billed per token, per minute or per 5 seconds show "up to", with the real amount settled when the task finishes; - When a model is unpriced or misconfigured, the estimate slot shows the matching notice rather than a wrong number. (Image: Parameters: aspect ratio, resolution, prompt optimization, output format, batch) ### Submit and results After you hit send, the node goes Queued, then Running, and the result lands back on the node when it finishes. - You can cancel mid-run: cancelling does not refund spent credits, and the node reverts to its last successful result; - On failure the system decides automatically whether to refund — only a "Refunded N credits" badge on the node means a refund actually happened; - Double-clicking send never double-charges; there is concurrency dedupe protection. > The quote and the real charge share one basis — what you see is what you pay · "Up to N credits" means metered billing, settled on completion · Cancel doesn't refund; failure may refund by the rules ## Text nodes A text node holds words: your own ideas and prompts, or text produced or reverse-engineered by AI. It often serves as upstream input for image and video generation. ### What they are for Text plays two roles on the canvas: as a prompt or script wired into a downstream image / video node to shape its generation; and as the product of a text-generation task in its own right (AI continuation or rewriting). ### Create and generate - From the Add node menu choose Text to create a text-generation node, then type or let the AI continue / rewrite; - Wire a text node's output into an image / video node and its words merge with that node's own prompt to shape the run; - What you type by hand in a text node is never machine-translated — it stays exactly as written. (Image: A text node on the canvas, usable as an upstream prompt for image / video generation) ### Prompt from image and speech to text Two more node types also produce text, created from a neighbor's "+" under the Text category: - Prompt from image: reverse-engineer a prompt that describes a picture, handy for reuse or tweaking; - Speech to text: extract words from a piece of audio. Both are paid generation, and you see the credit estimate when you create them too. > Text can be an upstream prompt or a generated product · Hand-typed text is never machine-translated · Prompt-from-image and speech-to-text also land as text nodes ## Image nodes Image nodes generate, hold and edit still frames, usually the first visual step before video. Select an image node and a toolbar appears above it, gathering the common image-editing tools in one place. ### Creating an image node - From the Add node menu choose Image — text to image by default; - Connect an image in and it switches to image-to-image, editing on top of that picture; - Or upload / drag / paste an image directly as an asset node for later generation. (Image: Image node with the generator bar: text-to-image by default, image-to-image once a picture is connected) ### Image editing tools With an image node selected, the main toolbar tools (with billing noted): - Crop: drag handles to crop, preset ratios or custom (free); - Multi-angle: drag a 3D cube to pick a new camera angle and tilt / scale, generating other views of the same scene (paid); - Repaint: brush or box the area to change, describe the change, and AI regenerates that region (paid); - Relight: drag light balls to adjust key / rim direction, intensity and color temperature, regenerating the lighting (paid); - Panorama: generate the image as a 2:1 panoramic frame (paid); - Rotate & mirror: preset or arbitrary rotation, horizontal / vertical mirror (free); - More → Outpaint: drag the frame outward to extend the picture, pick ratio / resolution, generate (paid); - More → Erase: box out what to remove and generate, no extra description needed (paid); - More → Markup: brush / box colored marks on the image and save as a new picture (free); - More → Photo enhance: scale up sharpness proportionally, or AI-repaint finer detail (paid); - More → Resize: enter a target width/height (aspect lockable) to resize directly (free); - More → Remove background: keep the subject, drop the background (paid); - More → Quick split: cut the image into a 2×2 / 3×3 / 4×4 grid and beyond (free). (Image: Select an image node with a result and the toolbar shows crop, multi-angle, repaint, relight, panorama, rotate, viewer, more, save, download, fullscreen) ### Panorama viewer and image compare - Panorama viewer: with a result image, open a WebGL 720° spherical preview — drag to rotate, scroll to zoom — and capture 4 or 12 viewpoints at once back to the canvas as references (free); - Image compare: wire two images into one compare node and drag the middle slider to compare them (display only, no new asset, free). > The toolbar separates free tools from paid generation · Repaint / erase / outpaint need you to mark a region first · Compare and the panorama viewer are free, display-only tools ## Video nodes Video nodes generate and process motion. They generate straight from text, carry first/last-frame and reference-to-video modes, and come with a full set of post-processing tools. ### Creating a video node - From the Add node menu choose Video — text to video by default; - Connect one image for image-to-video; two for first/last-frame; reference-to-video and other modes are supported too; - From the "+" beside an image node, expand the Video category to create an image-to-video node with the connection made for you. (Image: Video node: text-to-video, image-to-video, first/last-frame and reference-to-video) ### Video editing tools With a video node selected, the main toolbar tools (with billing noted): - Video upscale: scale to a higher resolution (720p / 1080p / 2K / 4K, billed per minute); - Trim: drag in / out points to cut a segment (free); - Video analysis: produce a structured shot breakdown (shot size, camera moves, etc.) (paid); - Video editor: open the multi-track timeline editor (export free); - Split A/V: pull out a silent video and an audio track as two results (free); - Mix audio: pick another audio to mix or replace (free); - First/last frame capture: pull the first and last frame as two images (free); - Vocal separation: keep only vocals, or only the background (free); - Subtitle removal: pick subtitle type and quality to strip subtitles from the video (paid); - Video translation: pick source / target language and method (subtitle translation or AI dubbing) (paid). ### Lens combo In the video (and image) generator bar you can, like an old camera, use four wheels to pick body, lens, focal length and aperture — carried into generation as prompt fragments. It only affects the one run you fire next and is not billed separately. > Text-to-video, image-to-video, first/last-frame, reference-to-video · Upscale is billed per minute; trim / split / capture are free · Lens combo is a prompt-level parameter, not billed separately ## Audio nodes Audio nodes cover voice and sound: text to speech, voice cloning and design, and the audio-processing tools that go with video. ### Text to speech From the Add node menu open Audio → Speech to create a text-to-speech node: - Type the text to read aloud, pick a voice and parameters, and generate a clip; - The bar has rate, pitch and volume prosody sliders and pause / emphasis markup helpers (some are model extensions; a model that doesn't support them reads the markup as plain text); - Text length is capped at about 50,000 characters, a soft on-screen counter. (Image: Open Audio from the Add-node menu to create a speech node) ### Voice cloning and design - Voice cloning: upload a reference clip to register a custom voice you can then synthesize with; - Voice design: design a new voice from a description, without a reference clip; - Registering a voice charges a registration fee and runs asynchronously in the background; a voice that is Processing has to become Ready before you can use it. ### Related audio tools Many audio capabilities actually live on the video node's toolbar (see Video nodes): split A/V, mix audio, vocal separation and the like all pull sound out or replace it, mostly for free. > TTS has prosody sliders and pause/emphasis markup · Cloning / designing a voice has a fee and becomes ready async · Split/mix and other audio tools are on the video node toolbar ## Connections and workflows Wire nodes together and one node's output flows into the next node's input — that is a workflow. This chapter covers connection rules and how to tidy up, group and run a whole group. ### Connection rules - Connect: drag from one node's output port to another node's input port; - Single-slot replace: most input ports take one connection at a time; a new one auto-replaces the old (to swap a reference image, just drag a new connection — no need to disconnect first); - Disconnect: right-click the node or use Disconnect all in the toolbar to remove every connection at once. (Image: Connected nodes: one node's output flows into the next node's input) ### Tidy up and group - Auto-layout: ⇧⌥F (Shift+Alt+F on Windows) rearranges nodes so connections read cleaner; you can also pick Optimize workflow layout from the right-click menu; - Group: marquee-select several nodes and press ⌘/Ctrl + G; a toolbar appears above the group frame; - Ungroup: ⌘/Ctrl + Shift + G; Batch download: the group toolbar can download all results at once. ### Run group and storyboard groups - Run group: runs every generatable node in the group in connection order, driving the whole chain in one click; - Convert to storyboard group: available only when the group is all image nodes (⌘/Ctrl + Alt + G); the toolbar then switches to ratio, grid count, index toggle, resolution, and Stitch (combine the group's images into one big picture). ### Save as a reusable workflow After grouping some nodes, the group toolbar's Add to toolbox saves the group as a reusable workflow (node structure and config only, not the actual results). Find it later under the left toolbar's Toolbox → My workflows and click Send to canvas to reinsert it. The Official templates tab in the Toolbox holds the product's ready-made workflows. > Input ports are single-slot — a new connection replaces the old · ⌘/Ctrl+G groups; then run the group or convert to a storyboard group · Add to toolbox saves node structure as a reusable workflow ## References and @ mentions In BLOCKLORE, there are two ways one node takes another's material: a connection (a hard reference), and an @ mention in the prompt. ### A connection is a hard reference Wire an upstream node into the current one and its output (image, video, audio, text) becomes an input to the current run. A connection is a "hard" reference — it states exactly where the data comes from and goes to. Remove the connection and the reference is gone. (Image: A selected node with edges: a connection is a hard reference) ### @ mentions to reference assets Type @ in a node's prompt box and an asset picker opens, split into Connected nodes and My library. Pick one and a removable reference chip is inserted into the prompt, submitted along with the run. This lets you bring a past asset into a run without wiring a connection. ### First/last frame and reference order When a node takes several reference images (say a video's first and last frame), the order is the meaning itself: the order among the connections decides which is the first frame and which is the last. Reorder to change the meaning — no separate role setting needed. > A connection is a hard reference; @ is a prompt-level one · @ can reference connected nodes or your saved library · First/last frame is told apart by connection order ## Video editor The Video editor node is a multi-track timeline editing container; opening it launches a standalone editor for stitching, layering and exporting clips into a finished cut. ### Opening the timeline editor Add a Video editor node from the Add node menu (marked Beta), or select a video node and click Video editor in the toolbar. Wire in the clips to compose, open the node, and you enter the full-screen timeline editor. (Image: Add the Video editor node (marked Beta) from the menu; it opens a multi-track timeline editor) ### Multi-track editing The editor is a multi-track timeline; common actions (shortcuts work inside the editor): - Play / pause preview: Space; - Split the clip at the playhead: S; - Delete the selected clip: Delete or Backspace; - Undo / redo: ⌘/Ctrl + Z, ⌘/Ctrl + Shift + Z; - Drag a clip to move, snap and swap with a neighbor; drag a clip's edge to trim its length; - Close the current pop-up menu: Esc. ### Exporting The export dropdown at the top-right of the editor has two options: - Export to local: auto-downloads to your computer when the compose finishes; - Export to canvas: appears as a new node on the canvas when the compose finishes. If the timeline has a gap with no clip covering it, you only get a hint — it will not block export. > Video editor is a full-screen multi-track timeline editor · S splits, Space previews, Delete removes a clip · Export to local or to canvas; a gap won't block export ## Director Stage The Director Stage node opens a 3D stage. You place cameras and virtual characters, adjust the framing, then screenshot to freeze the shot as an image to use as a reference for later generation. ### What the 3D stage is for When you need a precise, reproducible composition instead of leaving the camera to luck, the Director Stage helps: set up the camera and subject in 3D first, then screenshot that composition as a reference to feed generation. (Image: Add the Director Stage node (marked NEW) from the menu to place cameras in a 3D stage) ### Basic controls - Drag objects in the scene to reposition them; - X toggles grid snapping; - T (top) / Y (front) / Q (reset) switch preset views quickly; - Group / ungroup: ⌘/Ctrl + G, ⌘/Ctrl + Shift + D, to manage multiple objects in the scene; - Delete the selected object: Delete or Backspace. ### Placement and screenshot - The stage content saves into the node config automatically — no manual save; - The screenshot button turns the current view into a real image asset that appears on the canvas, connected to the Director Stage node so you can trace where it came from. > The Director Stage is for precise, reproducible framing · T/Y/Q switch views, X toggles grid snap · A screenshot becomes a real image asset, connected for lineage ## Assets and history Everything you generate, upload and favorite on the canvas goes into the asset library (/assets). It is where you view, organize and download your history. ### History The History tab gathers all your generated output — filter by image / video / audio / document, sort, favorite, download, delete. It records what you have generated, not the contents of any one canvas. (Image: Assets: History filters by image / video / audio / document, with sort, favorite and download) ### Saved Saved is the collection you have favorited — view by category, rename, recategorize, download, delete — handy for keeping frequently used assets together. ### On putting assets back and batch download - These two standalone pages only view, favorite, download and delete; they do not put an asset back onto a canvas — the page has no way to know which canvas you mean. To reuse a past asset on the current canvas, go back to the editor and pick it via Choose from generation history in the Add node menu; - Batch download fires a browser download per selected item; packaging into a single zip is not supported yet. (Image: In the editor's Add-node menu, Choose from generation history puts a past asset back onto the current canvas) > History = everything you've generated; filter/favorite/download · Put assets back via Choose from generation history in the editor · Batch download is one-by-one; no zip packaging yet ## Publishing to Community Once a canvas is ready you can share it read-only for others to view, or publish it to the Community showcase. This chapter covers both. ### Share a read-only link Click Share in the canvas top bar to generate a read-only link (/shared/a-token). The Access setting in the dialog toggles: - Public: anyone with the link can view the canvas (no sign-in needed), but only view, not edit; - Only me: turn Public off and the link stops working for others. Sharing only grants read-only viewing, never edit access. (Image: The Share and Publish entries in the canvas top bar) ### Publish to Community Click Publish to send a work to the Community showcase; you fill in: - A cover image and/or video (at least one; upload locally or pick from your generation history); - Title and description (required); - Activity tag, contest track, social link (optional); - The Public canvas switch: on (default) lets others click Fork project on your work's detail page to copy the canvas and keep creating; off keeps the work on display but blocks copying; - The Open-source canvas switch: on additionally publishes a forkable open-source snapshot, with a license of your choice; - The compliance authorization checkbox (required). (Image: The Community showcase where published works appear) ### View the process and fork After publishing, anyone (including signed-out visitors) can click View process on the work's detail page to browse the full node structure read-only, no sign-in required; only clicking Fork project to copy it into their own account for editing requires signing in. > Share = read-only link; Publish = into the Community showcase · Publishing requires a cover, title, description and compliance consent · The Public canvas switch decides whether others can fork your canvas ## Creator contests BLOCKLORE runs official creator contests from time to time, rewarding strong work with credits and a homepage feature. This chapter covers finding and entering them. ### Where to find contests Entry points for ongoing contests usually appear on the Blog page and in the AI Plays area as tagged image cards; open a contest's detail page to see its theme, dates, rules and reward tiers. (Image: Blog page: entries for ongoing contests show as tagged cards) ### How to enter Entering is simply "attach a contest track when you publish": pick an ongoing contest under Contest track in the Publish dialog, submit, and your work links to that contest (the publish flow is in Publishing to Community). ### Rewards and features A contest generally has a credit prize pool and homepage feature slots; the rules, deadline and reward tiers are whatever the contest's detail page states. > Contest entries are on the Blog page / AI Plays area · To enter, pick the contest under Contest track when publishing · Rules and rewards are whatever the detail page states ## Credits and membership Generation is billed, in credits. This chapter covers how credits are used, how to top up, and the refund rules. ### How credits are used - Before each paid generation, the bar shows the credits this run will cost; you confirm before submitting; - Models differ in capability and price; those billed per token / minute / 5 seconds show "up to N credits", with the real amount settled on completion; - Your balance shows live in the top bar and the account menu. ### Membership and top-up packs - Membership plans grant monthly credits per cycle; top-up packs are one-off credit purchases you make as needed; - The Pricing / membership page (/pricing) lists the plans and packs and can switch display currency (display conversion only; settlement stays on the real basis); - Teams have their own shared credit pool; generation in team projects draws from it. (Image: Pricing / membership: plan tiers, monthly credits and the volume you can generate) ### Refund rules - Cancelling an in-progress generation does not refund spent credits; the node reverts to its last successful result; - On failure the system decides automatically whether to refund — only a "Refunded N credits" badge on the node means a refund actually happened; no badge means no refund. > Check the estimate first; "up to N" means metered billing · Membership grants credits per cycle; packs are one-off · Cancel doesn't refund; on failure watch for the refund badge ## Teams and settings If you want to create with others on the same canvas, you need a team. This chapter covers creating a team, roles and permissions, and real-time collaboration. ### Create a team and share credits - Go to the Teams page (signed-in, no special permission), click Create team, fill in a name and description; the creator becomes the owner automatically; - A team has a shared credit pool; generation in team projects draws from it, while your personal projects still use your own credits; - To put a canvas under a team: on the Workspace page switch to the Team tab, pick a team you have rights in, and click New canvas. (Image: Teams page: create a team, share a credit pool and collaborate with members) ### Roles and permissions Four levels, low to high: - Viewer: can only view team canvases, not edit, and won't see rename / duplicate entries; - Editor and up: can actually edit team canvas content; - Admin and up: can invite / remove members, change member roles, delete team canvases, and create canvases under the Team tab; - Owner: additionally can transfer ownership and disband the team; non-owners can Leave team anytime, while an owner must transfer or disband first. Inviting members: on the team detail page, admins and up can add by email directly or generate an invite link to share, specifying the invitee's role. ### Real-time collaboration cues On a team canvas: others' cursors show live with name labels; when someone selects a node it gets a colored outline labeled "So-and-so is editing"; the top shows "N online". These are display-only cues and never block you from editing and saving. > A team has a shared pool; team projects draw from it · Four roles: viewer < editor < admin < owner · Collaboration cursors/outlines are hints, not edit locks ## Tips and FAQ This last chapter rounds up common shortcuts, what the save states mean, and a few frequent questions. BLOCKLORE detects your system automatically: Mac shows ⌘ ⌥ ⇧, while Windows / Linux convert to Ctrl / Alt / Shift. ### Common shortcuts - Generate (a selected generatable node): ⌘/Ctrl + Enter; - Open the Add node menu: Tab; - Delete selection: Delete or Backspace (won't fire while the cursor is in an input, to avoid accidental deletes); - Undo / redo: ⌘/Ctrl + Z, ⌘/Ctrl + Shift + Z (up to 50 undo steps); - Copy node to the in-app clipboard: ⌘/Ctrl + B; Duplicate: ⌘/Ctrl + D; - Copy image to the system clipboard: ⌘/Ctrl + C (with an image node that has a result selected); - Group / ungroup: ⌘/Ctrl + G, ⌘/Ctrl + Shift + G; Auto-layout: ⇧⌥F; - Zoom: ⌘/Ctrl + = / -, Fit view: ⌘/Ctrl + 0; Marquee / hand: V / H. The Shortcuts button on the left toolbar opens a cheat-sheet panel anytime, where you can toggle between the Mac and Windows key styles. (Image: The Shortcuts button on the left toolbar opens the cheat-sheet panel anytime) ### What the save states mean The canvas has no manual save button; the top bar shows status live: - Saved: the current content is saved; - Saving: edits are being written to the server; Unsaved: there are edits not yet submitted (usually auto-turns to Saving within seconds); - Save failed, retrying: a network issue caused a failure; the system keeps retrying on its own until it succeeds, no manual action needed; as long as you don't close the whole browser app, edits from this window won't be lost; - Version conflict: rare; two buttons appear — Copy local draft and Reload server version — better to copy the local draft as a backup first, then reload; - Read-only: you have no edit rights right now (e.g. viewing someone's read-only shared link). ### Frequently asked - Does cancelling refund credits? No; the canvas reverts to the last successful result; on failure, watch for the "Refunded N credits" badge; - Can't find the Script node in the menu? It is only created from the Story script quick-start card on an empty canvas; - Will edits be lost if I close the browser before a save lands? No; next time you open the canvas it auto-restores edits that didn't sync during the outage; - Can a read-only share let others edit? No; to co-edit, put the canvas in a team project and invite them as Editor or above. > ⌘/Ctrl+Enter generates, Tab adds a node, ⌘/Ctrl+0 fits · "Save failed, retrying" recovers on its own — no need to act · The Script node is via the quick-start card; cancel doesn't refund # Blog ## Meet the AI Agent That Runs Your Production Canvas Published 2026-08-28 by BLOCKLORE Team. The bottleneck in AI-assisted content creation has rarely been the generative models themselves. It's the orchestration: briefing the concept, keeping every shot visually consistent, catching problems before they derail a deliverable, and finally assembling everything into a finished cut. That's usually where a human producer spends most of their time. BLOCKLORE's canvas-based AI agent is built to take on that producer role directly, working alongside you inside a visual node canvas rather than a simple chat window. ### A creative partner, not just a prompt box The agent starts by asking the right questions. Instead of generating blindly from a vague brief like "let's make an epic drama," it asks for a core concept or story seed, and lets you either pick a direction or describe your own. From there, it builds out a structured plan, breaking a single idea into a sequence of shots. Each shot becomes its own node on the canvas, complete with prompts, generated keyframe images, and matching video clips. ### Multi-modal by design The canvas isn't limited to images. In the same workspace, the agent generates and organizes video clips, a musical score, and ambient sound effects, laying each out as a connected node ready to feed into a final assembly. Because everything lives on one canvas, you can see the entire production - visuals, motion, and audio - as a single interconnected pipeline rather than juggling separate tools. ### Iteration through natural conversation What stands out most is how the agent handles revision requests in plain language. Ask it to add a transitional shot because a cut feels too abrupt, or to make a lighting effect cascade across every background element in a scene, and it translates that creative note directly into new generation nodes, placed and wired into the pipeline, and queued for your confirmation. It doesn't just generate in isolation; it explains what changed and why, and tells you what to expect once you approve the run. ### Keeping the story consistent One of the harder problems in AI-generated sequences is visual consistency: the same character, environment, and style holding together from shot to shot. When asked to address this, the agent rebuilt the shot set with consistency as the explicit goal, then reconnected the entire pipeline so every video and audio node fed correctly into a unified final-cut assembly. ### Cleaning up after itself Long creative sessions generate a lot of scaffolding: early drafts, discarded takes, one-off experiments. Rather than leaving that clutter behind, the agent can audit the whole canvas on request, remove nodes that are no longer part of the active pipeline, and re-wire the surviving elements so the workspace reflects only the production actually being used. It also reports back with a clear summary of what was deleted and what was reconnected, so nothing changes without visibility into the reasoning. ### Solving problems along the way Creative pipelines don't always run smoothly on the first pass, and the agent doesn't just report failures back to the user. It's capable of diagnosing and resolving issues that come up during the creation process itself, then re-queuing the affected work with a clear explanation of what was adjusted and why. ### Producer-level thinking, on your canvas Taken together, the workflow looks less like prompting a model and more like directing a small production team: brief the concept, review generated assets, request changes in everyday language, and let the agent handle sequencing, consistency, and assembly. That shift, from single-shot generation to an agent that manages an entire visual production end-to-end, is the capability worth paying attention to. [Video](https://blocklords.ai/brand/blog-canvas-agent.mp4) ## Seedance 2.5 Best Practices: A Beginner-Friendly Guide Published 2026-08-18 by BLOCKLORE Team. Seedance 2.5 is ByteDance's latest AI video model. It can generate up to 30 seconds of video in one pass, with sound built in at the same time, and it lets you feed it reference images, clips, and audio so it sticks to a specific character, style, or camera move. That's a lot of power, and it's easy to either underuse it (typing one vague line and hoping) or overload it (dumping in 40 references and a giant paragraph). Here's how to get good, consistent results on BLOCKLORE without needing a professional production workflow. ### 1. Describe one shot at a time Write your prompt the way you'd describe a single shot to a camera operator: who's in it, what they do, where they are, and how the camera moves. Avoid cramming multiple scenes or ideas into one prompt; the model handles a focused shot much better than a jumbled one. > Example: "Medium shot of a woman in a yellow raincoat walking down a rainy city street at night. She stops, looks up at a neon sign, and smiles. Camera slowly tracks alongside her." ### 2. Use a handful of references, tagged clearly Seedance 2.5 accepts references (images, short video clips, audio) and lets you tag each one with @Image1, @Video1, @Audio1, and so on in the prompt text. You don't need to fill every slot - 3 to 5 well-chosen references (a face, an outfit, a location) usually give more stable results than a large, unlabeled pile. The important part is naming what each one is for, since the model won't reliably guess on its own. > "@Image1 shows the character's face and jacket, keep these consistent. @Image2 is the coffee shop interior, use it for the setting." > "Use @Image1 for the character's face. Use @Image2 for his leather jacket. Use @Image3 for the lighting and color grade." ### 3. Don't default to the full 30 seconds A longer clip sounds appealing, but the more time the model has to fill, the more chances there are for details to drift - a face changing slightly, an object disappearing. Start with a shorter clip (5-10 seconds) for anything with a lot of detail or a specific look you want to protect, and only reach for the full 30 seconds when the action genuinely needs that much room, like a scene with a clear beginning, middle, and end. A 30-second video is a long time for an AI to guess what you want. If your prompt is just "A man walks down the street," the model might have him walk for five seconds and then start doing backflips. > Example prompt: "Use @Image1 for the main character. 16:9 widescreen, smooth cinematic tracking shot. 0-10s: The character walks down a dark, rainy alleyway, illuminated by neon signs. 11-20s: He stops, turns toward the camera, and pulls a map from his pocket. 21-30s: The camera slowly pushes in for a tight close-up on his face as he looks confused. Audio: raindrops, heavy footsteps, no music. Strictly preserve the face from @Image1." ### 4. Write for the sound, not just the picture Because audio is generated together with the video, describe what you want to hear, not just what you want to see. If you leave sound out of the prompt, you're leaving it up to chance. Mention dialogue, ambient noise, music, or explicitly say "no music" if you want a quiet, natural scene. > "Audio: rain on the pavement, distant traffic, footsteps on wet concrete, no music." > "Sync the camera movements to the rhythm in @Audio1. When the heavy bass drops at 0:15, execute a fast whip-pan to the right." ### 5. Build long scenes by extending, not by overloading one prompt If you're trying to tell a longer story, don't try to script all of it into a single request. Generate a first clip, then use it as the basis for an extension - a follow-up generation that continues from the previous one, keeping the same characters and setting. This keeps each individual prompt focused and gives you a checkpoint to review before continuing. > Extension prompt: "Continue from the previous clip. Same character, same street, same visual style. She now crosses the street and enters the coffee shop." ### 6. Fix small mistakes with editing, not a full re-roll If a clip is mostly right but one detail is off - a wrong prop, a background object, a distorted hand - you don't need to regenerate the whole thing from scratch. Seedance 2.5 supports edits that target a specific part of an existing clip while leaving the rest untouched. In your prompt, reference the original clip, explicitly say what should stay the same, and describe only the change you want. > "Edit the previous video. Keep the character, camera movement, and background unchanged. Replace only the cup on the table with a glass of water." For time-specific fixes, you can also break the clip into time ranges in your prompt (like "0-5s" and "5-10s") and describe what should happen or change in each range, rather than editing the whole clip at once. > "Keep 0-20s completely unchanged. From 21-30s, change the character's action so he looks at his phone instead of the sky." ### 7. Save what works When a generation nails a character's look or a scene's lighting, save that still or clip. You can reuse it as a reference in future prompts to keep things consistent across a project, instead of re-describing everything from scratch each time. ### Quick recap Keep prompts focused on one shot, use a few well-labeled references instead of many, don't over-use the max clip length, describe your audio, extend rather than cram, edit instead of regenerating, and save your best results as future references. Like most AI tools, Seedance 2.5 rewards iteration - expect to refine a shot over a couple of tries rather than nailing it on the first attempt. [Video](https://blocklords.ai/brand/blog-seedance-25-best-practices.mp4) ## Seedance 2.5 Launch: New Models, UI Improvements, and Discounts Published 2026-08-12 by BLOCKLORE Team. As our platform continues to grow, we are expanding our capabilities with a suite of powerful new AI models, improved workflow features, and special discounts. ### The Centerpiece: Seedance 2.5 Seedance 2.5 marks a major step up from Seedance 2.0, focusing on long-form storytelling, multimodal reference, and more advanced editing tools. It was officially unveiled on June 23, 2026, at ByteDance's Volcano Engine FORCE conference, and is being called one of the biggest single jumps in AI video generation to date. This video generation model is a massive leap forward for digital creators, filmmakers, and marketers. Seedance 2.5 introduces native 30-second one-shot video generation at stunning 4K resolution. Unlike older models that require stitching multiple short clips together, Seedance 2.5 maintains strict spatial and character consistency from the first frame to the last. What truly sets this version apart is its advanced multimodal reference system. You can now upload up to 50 distinct reference inputs per generation. By combining character images, environment videos, and audio clips, you get granular control over exactly what appears on screen. Furthermore, Seedance 2.5 introduces precise region-level editing. If a single detail is incorrect, you no longer have to regenerate the entire video. You can isolate the specific area, and the model will re-render just that region without losing the pacing, facial expressions, or lighting of the original take. Coupled with a 20% improvement in prompt adherence and native audio synchronization, Seedance 2.5 brings production-ready control directly to your workspace. ![Seedance 2.5 live on the BLOCKLORE homepage](https://blocklords.ai/brand/blog-seedance-25-ui-1.jpg) ### Massive Model Library Expansion To ensure you have the best tools for every task, we have heavily expanded our offerings across all creative categories: Video Generation: Alongside Seedance 2.5, we have integrated Wan 3.0, expanding your options for fluid dynamic motion and creative visual generation. Image Generation: Step up your visual fidelity with the brand-new Qwen-Image 3.0 Pro, or explore unique artistic stylings with the newly added Grok Imagine. Text and Reasoning: Our text capabilities have received a massive boost. We are proud to introduce OpenAI's latest GPT 5.6 lineup (featuring the Sol, Terra, and Luna tiers) as well as GPT 5.5. We have also integrated the powerful Qwen 3.8 Max, Dola Seed 2.1 Turbo, and the newest models from xAI: Grok 4.5 and Grok 4.3. ![The new AI Video panel: one line, straight to video](https://blocklords.ai/brand/blog-seedance-25-ui-2.jpg) ### UI Enhancements & Bug Fixes To accommodate these powerful new tools, we have deployed a major Homepage UI/UX overhaul for smoother navigation and a cleaner aesthetic. We have also introduced a dedicated AI Video and AI Image Panel for quick generation, drastically reducing the friction between writing a prompt and receiving your media. Finally, we have implemented several under-the-hood bug fixes and performance optimizations to keep the platform running seamlessly. ### Special Discounts To celebrate this expansion, we are rolling out limited-time promotional pricing. Enjoy special discounts on our Seedance 2.0 Fast and Seedance 2.0 Mini models starting today. Also, for a limited period of time, certain credit packs will be heavily discounted, and users can find these offers in the credit store section. ![Limited-time offers in the credit store](https://blocklords.ai/brand/blog-seedance-25-ui-3.jpg) [Explore plans & discounts](https://blocklords.ai/en/pricing) ## Evolving the Arena: From BLOCKLORDS to BLOCKLORE Published 2026-07-03 by BLOCKLORE Team. The landscape of digital ownership and creator economies is moving at a rapid pace. Over the last few years, the Web3 gaming sector has undergone a massive structural shift, moving past speculative systems toward sustainable utility. In this consolidating market, projects are discovering that blockchain technology serves as a creative substrate capable of unlocking entirely new paradigms of digital expression. Today, a monumental step forward is being taken in this journey. We are evolving our focus toward a broader, highly scalable vision: BLOCKLORDS has officially evolved into BLOCKLORE, a node-based, collaborative AI generative canvas designed to change how media is created, connected, and shared. This transition does not represent an exit from Web3; it is a strategic expansion. By moving the infrastructure from a single gaming engine to a decentralized creative platform, the brand is positioned at the forefront of the generative AI landscape. ### What is BLOCKLORE? Traditional creation tools keep prompt engineering, visual generation, and post-production trapped in isolated silos. BLOCKLORE breaks down these barriers by introducing an interactive, node-based workspace where you can build, drag, and link complex multi-agent generative workflows. BLOCKLORE delivers a complete AI media canvas with a full-process generative workflow, from initial layout and concept draft to high-fidelity video editing. Powered by top-of-the-line machine learning models, it lets digital artists, storytellers, and prompt engineers chain operations seamlessly: link a character design node to an environment template, route it through an animation model, and generate a cohesive, cinematic sequence in minutes. Most importantly, the platform is committed to making advanced creation accessible. By using highly optimized compute pipelines, BLOCKLORE offers these powerful creative tools at highly competitive prices, lowering the infrastructure barriers that historically blocked independent creators. ### Honoring the core base: the $LRDS loyalty airdrop A great brand relies on the community that stood by it first. For three years, active players and supporters built, fought, and contributed to the legacy of the game. To recognize that historical commitment, a loyalty airdrop of 3,000,000 $LRDS tokens is launching soon. This distribution is dedicated strictly to rewarding the core active players and supporters who laid the foundation for the ecosystem. The airdrop reflects real engagement with the BLOCKLORDS world: how long players have been part of the ecosystem, the assets they collected, and their overall game and ecosystem activity — rewarding the people who helped build this world with us. Rather than automating it through complex, developer-heavy smart contracts, which would take precious engineering hours away from perfecting the canvas, the airdrop will be managed manually and transparently. ### Direct connections for the $LRDS token The native token remains a central part of this ecosystem. $LRDS will have direct links to BLOCKLORE and to the legacy of BLOCKLORDS very soon, bridging the historical achievements of the player base with the next phase of the platform's evolution. $LRDS can still be staked on the Lordchain website. ### The creative era begins now By combining Web3 infrastructure with next-generation generative AI, BLOCKLORE is building a playground where creative focus is treated as the most valuable asset in the ecosystem. Prepare the concepts, and get ready to link ideas: the canvas is officially open, and the future of media creation is ready to be written. [Open the canvas](https://blocklords.ai/en/dashboard) ## Mastering the Canvas: A Step-by-Step Guide to Creative Workflows Published 2026-07-03 by BLOCKLORE Team. The digital art landscape has entered an era of rapid transformation, where prompt engineering, visual generation, and video post-production are converging into a unified space. Traditional creation platforms keep these processes separated, forcing artists to juggle disconnected browser tabs, download temporary files, and rebuild prompts from scratch. BLOCKLORE resolves this operational friction with the Canvas: a node-based, collaborative generative environment that changes how complex visual stories are built, connected, and shared. In a single flexible workspace, you manage every phase of production, from the first concept draft to frame-accurate video editing. ### 1. The core architecture of the canvas Unlike traditional, linear software interfaces, BLOCKLORE runs on a single pannable and zoomable design surface. This infinite workspace functions as a visual database where an entire project stays organized in one place. You add visual blocks, known as nodes, from a menu; connecting the output port of one node to the input port of another establishes a logical relationship that passes generation results straight to the next step. Every intermediate result is retained permanently on the workspace as an independent node. Because these steps are never lost, you can branch creative paths, reuse earlier outputs, and compare variations side by side without overwriting historical work. Every generation result is also converted into a permanent asset and uploaded to private asset storage, so your work never depends on a temporary provider link or a local file that breaks over time. ### 2. The four core node types Text nodes hold prompts, narrative descriptions, and written content; these prompt variables are the foundational parameters that feed downstream generation blocks. Image nodes generate, upload, or edit static images, and their output can be reused anywhere on the canvas as a style reference, an image-to-image seed, or a direct visual component. Video nodes generate and process high-fidelity clips, using advanced motion models to turn static assets into dynamic sequences. Audio nodes generate or attach voiceovers, atmospheric sound design, and background music, synced to the visual layers. Beyond the standard generation blocks, the canvas offers specialized nodes for storytelling and post-production. The Script node acts as the logical director of the workspace, organizing storyboards, camera directions, and shot lists that automatically propagate prompt variables downstream. The integrated video editor timeline then turns those raw video nodes into a finished film. ### 3. Building an automated generative workflow Building a workflow means connecting output ports to the input ports of the nodes that follow. To generate a consistent character animation, a Script node feeds a Text node that defines the character's narrative description; that Text node routes into an Image node to render a consistent portrait; the portrait then drives a Video node to produce a fully animated sequence. Running a single node, or an entire chain, propagates data downstream automatically. This connection model enables complex, multi-agent workflows in which upstream visual parameters continuously guide and refine downstream outputs. ### 4. Accelerating design with automated templates BLOCKLORE ships with a suite of automated templates that take you from an abstract idea to a structured visual result. These packaged configurations assemble the required nodes and connections for you: consistent character sheets with multiple angles, multi-camera grids for varying scene perspectives, and story progression frames that map out a narrative sequence with minimal manual setup. ### 5. Achieving precision with the Director Stage To remove the guesswork commonly associated with generative AI, the Director Stage gives you frame-accurate control over camera movement and subject positioning. You pose a 3D scene, define specific camera paths such as pans, tilts, or dolly zooms, and apply that spatial direction directly to the generative nodes, so the final output follows strict, repeatable framing rules instead of leaving composition to chance. ### 6. Managing generation and credit consumption Running generations consumes computing credits, and every model shows its cost on the canvas before it runs. Monthly membership plans grant a set allocation of credits per cycle, and top-up packs add computing allocations for heavy creation cycles. Because computational fees cannot be refunded once consumed, the canvas keeps costs transparent up front, showing the credit cost of every model run before execution. ### 7. Forking and community contests Once a visual narrative is finished, you can publish the work and its canvas to the public community showcase, or download it locally. Because every published piece stays natively linked to the canvas that generated it, other creators can explore, study, and even copy the original template. Forking copies a successful node layout straight into another workspace, where new text nodes or visual elements build a brand-new workflow. This collaborative architecture is the core of the ongoing BLOCKLORE creator contests, which regularly distribute cash, credits, and visibility to reward outstanding work. Sign in, choose a canvas workflow, and turn a simple concept draft into a fully realized cinematic narrative. [Open the canvas](https://blocklords.ai/en/dashboard) ## Video Editor Guide: Navigating the Timeline and Composing Final Motion Assets Published 2026-07-03 by BLOCKLORE Team. The transition from individual generative nodes on the infinite canvas to a polished, unified motion asset happens inside the BLOCKLORE video editor. This interface is a structured, non-linear editing space where you arrange, trim, split, apply transitions, and configure multiple video generations and audio tracks into a single high-fidelity film. This guide walks through the core functional zones, the editing sidebar, the timeline controls, and the technical parameters that shape the final composition. ### 1. The creative sidebar and asset library The left-hand sidebar is the primary navigation and asset hub, organized into five specialized tabs. The Media tab opens the scrollable generation library, displaying native generations, video-to-video renders, and imported clips as visual thumbnails. The Transition tab holds visual effects that smooth the cut between clips. The Text tab overlays, styles, and positions text directly on the video frames. The Captions tab controls automated subtitle generation and styling. The Audio tab manages sound effects, background tracks, and voiceovers. ### 2. Timeline ingestion and multi-track assembly The editor is built for a clean drag-and-drop or click-to-add workflow. Click the "+ Add media" block at the start of the primary timeline track to import assets from the media library, or drag clips straight from the sidebar or the infinite canvas onto the timeline. Directly beneath the primary media track, a secondary "+ Add audio track" option lets you place, align, and sync independent background music, sound effects, or voiceover layers. The timeline ruler marks precise chronological intervals starting at second zero, so exact durations stay readable as clips are placed. ### 3. Playback, navigation, and zoom controls The central preview monitor renders the composition in real time, with a dynamic playhead that tracks across the timeline and corresponds directly to the frame on screen. The zoom slider on the middle-right adjusts the horizontal scale of the timeline: zoom in for frame-by-frame adjustments, zoom out for a bird's-eye view of the whole sequence. Next to it, the editor shows the playhead's position against the total length of the composition, alongside a percentage scale indicator. ### 4. Timeline editing and precision tools A row of one-click tools sits directly above the timeline track. Play and pause controls sit beside undo and redo arrows for reverting or re-applying sequence edits. The split tool, marked by a scissor icon, slices the active clip at the exact position of the playhead to isolate segments, delete unwanted frames, or insert a secondary clip. "Trim left" discards every frame to the left of the playhead and "Trim right" discards everything to the right. The magnetic "Snap" toggle, highlighted when active, pulls adjacent clips together so no accidental gaps or black frames survive between scene transitions. The "Captions" toggle enables or disables subtitle tracks across the timeline, and the volume icon on the far left of the track mutes or adjusts the level of the attached audio layers. ### 5. Configuring global video parameters Before rendering, the technical output parameters have to match the target platform. They live in the "Video params" bar at the bottom of the interface. Resolution sets the master file output, such as 720p or the higher-definition presets available on your membership plan. Aspect ratio defines the frame dimensions, from 16:9 landscape for wide-screen films to vertical formats for mobile feeds. Frame rate controls playback smoothness and defaults to 30 frames per second. Fit mode, set to letterbox by default, decides how clips with different aspect ratios sit inside the uniform composition canvas, adding neutral bars instead of cropping the original proportions. ### 6. Exporting the completed composition Once clips are arranged, trimmed, and configured, the "Export film" button in the top-right corner compiles every linked clip, transition, text overlay, caption, and synced audio track into a single master file. The finished export is saved automatically to your secure, private cloud storage and is immediately available to download or share. [Open the video editor](https://blocklords.ai/en/dashboard) ## Visual Prompt Guide: Orchestrating Modular AI Generations on the Canvas Published 2026-07-03 by BLOCKLORE Team. Traditional generative AI tools limit creators to a single, text-heavy input box. Long paragraphs of descriptive text often lead to unpredictable results, because models struggle to weigh many instructions at once. BLOCKLORE redefines the process with visual prompting: decomposing a prompt into modular, interconnected nodes on an infinite canvas. By separating creative intent into individual, logical steps, visual prompting gives you absolute control over the generation pipeline. Here are the foundational techniques. ### 1. Modular prompting: breaking down the formula Visual prompting relies on breaking a complex scene into distinct components and assigning each one to its own Text node. Instead of writing a single prompt such as "a medieval warrior standing in a misty forest, highly detailed, cinematic lighting, slow pan," the elements are separated: a subject node defining only the character ("a medieval warrior in iron plate armor"), an environment node defining the setting ("a dense, misty pine forest at dawn"), and an aesthetic node defining style and lighting ("cinematic volumetric lighting, dark fantasy aesthetic, cool color palette"). Connecting these individual Text nodes to a single Image node lets the model process each instruction with optimal clarity, producing output that matches the creative vision far more closely. ### 2. Parent-child chains for style consistency Maintaining visual consistency across generations is one of the hardest problems in generative media, and parent-child node relationships solve it. When an Image node produces a successful character render, that node becomes the parent. To generate a different scene featuring the same character, link the parent's output to the input port of a downstream child Image node. The connection passes the original visual parameters downstream as a direct style reference. Attach a new Text node to the child to specify a new action, such as "holding a sword," and the model preserves the character's facial features and armor design while executing the new prompt. ### 3. Sequential workflows: from draft to animated video Visual prompting extends well beyond static images. Because the canvas also houses Video, Audio, and Script nodes, an entire production timeline can be mapped out visually in a left-to-right sequence: a Script node outlines the storyboard, camera moves, and shot list; it links to a Text node that generates specific prompt variables for each shot; the Text node connects to an Image node that renders the static scene layout; the resulting image is routed into a Video node where the frame is animated with precise motion controls; and an Audio node joins the final Video node to overlay synchronized sound design. Connecting node outputs to downstream inputs builds a directed acyclic graph that propagates prompting instructions automatically. Because every intermediate generation stays visible on the canvas as a persistent node, variations can be branched, reused, and compared side by side without losing progress. ### 4. Eliminating randomness with the Director Stage Traditional prompting leaves camera positioning entirely to the model. The Director Stage introduces spatial visual prompts that remove that randomness: pose a 3D scene, map out exact camera movements such as a dolly zoom or a slow cinematic pan, and feed those spatial parameters into the active prompt nodes as physical constraints. The model then renders according to those coordinates, delivering frame-accurate, repeatable compositions. With visual prompting, creators move from passive guesswork to active, structural orchestration. The infinite canvas turns every text prompt, style reference, and animation parameter into a tangible, connectable building block, making the whole process more organized, repeatable, and powerful. [Browse the templates](https://blocklords.ai/en/templates)